欢迎访问作物学报，今天是
作物学报

作物学报 ›› 2019, Vol. 45 ›› Issue (7): 1099-1110.doi: 10.3724/SP.J.1006.2019.81065

• 耕作栽培·生理生化 • 上一篇    下一篇

基于ESTARFM模型的区域农田高时空分辨率影像产生与应用

陈梦露1,2,李存军1,*(),官云兰2,周静平1,王道芸2,罗正乾3   

  1. 1 北京农业信息技术研究中心, 北京100097
    2 东华理工大学测绘工程学院, 江西南昌330013
    3 新疆农业科学院综合试验场, 新疆乌鲁木齐830091
  • 收稿日期:2018-09-17 接受日期:2019-01-19 出版日期:2019-07-12 网络出版日期:2019-02-27
  • 通讯作者: 李存军
  • 作者简介:E-mail: chenml712@163.com
  • 基金资助:
    本研究由国家自然科学基金项目(41671435)

Generation and application of high temporal and spatial resolution images of regional farmland based on ESTARFM model

CHEN Meng-Lu1,2,LI Cun-Jun1,*(),GUAN Yun-Lan2,ZHOU Jing-Ping1,WANG Dao-Yun2,LUO Zheng-Qian3   

  1. 1 Beijing Research Center for Information Technology in Agriculture, Beijing 100097, China
    2 East China University of Technology, Nanchang 330013, Jiangxi, China
    3 Xinjiang Academy of Agricultural Sciences Comprehensive Test Site, Urumqi 830091, Inner Mongolia, China
  • Received:2018-09-17 Accepted:2019-01-19 Published:2019-07-12 Published online:2019-02-27
  • Contact: Cun-Jun LI
  • Supported by:
    This study was supported by the National Natural Science Foundation of China(41671435)

RichHTML

40

PDF

736

摘要：

多时相遥感影像特别是关键生育期数据是农业物候、长势及产量监测的重要数据源, 然而可见光影像易受云雨干扰, 在特定区域关键时间窗口缺少高时空分辨率数据的现实情况下, 遥感影像时空数据融合方法变得尤为重要。增强型自适应反射率时空融合模型ESTARFM (Enhanced Spatial and Temporal Adaptive Reflectance Fusion Model)是一种小区域合成高时空分辨率影像的较好方法, 该算法在我国不同农业种植区的适应性及应用工作尚未充分展开。本文以河北、黑龙江、新疆典型农区为研究区域进行大面积应用检验分析, 基于MODIS与Landsat影像, 利用ESTARFM生成具有高时空特征的Landsat模拟影像, 将其与真实Landsat影像进行对比, 并在新疆地区展开ESTARFM算法在NDVI方面的应用。结果表明, ESTARFM对3个不同区域状况的地区都有较好的影像预测能力, 并且在新疆地区可以很好地生成30 m空间分辨率的多时相NDVI, 用于作物分类和长势监测。

关键词: 高时空分辨率, ESTARFM, 数据融合, NDVI, 长势监测

Abstract:

Multi-temporal remote sensing images are important data sources for agricultural phenology, growth, and yield monitoring. However, visible light images are vulnerable to cloud and rain, and there is a lack of high temporal and spatial resolution data in reality, the remote sensing image fusion methods have become particularly important. ESTARFM (Enhanced Spatial and Temporal Adaptive Reflectance Fusion Model) is used to synthesize high spatial-temporal resolution images in small areas. The adaptability and application of the algorithm in different agricultural growing areas in China have not yet fully developed. In this paper, the large area application test analysis was performed in the Hebei, Heilongjiang, and Xinjiang. Based on MODIS and Landsat images, we used ESTARFM to generate Landsat images with high spatial-temporal characteristics, which were compared with the real Landsat images. The application of ESTARFM algorithm in NDVI was performed for crop growth monitoring in Xinjiang. In conclusion ESTARFM can perform better image prediction in three different regional conditions, generate 30 m multi-temporal NDVI with good spatial resolution in Xinjiang, and monitor the growth of crops.

Key words: high spatiotemporal resolution, ESTARFM, fusion data, NDVI, growth monitoring

图1

研究区域地理位置"

表1

Landsat-8 OLI影像日期"

研究区域
Research area		 中心经纬度
Center latitude and longitude		 日期(年积日)
Date (DOY)
河北研究区 Hebei research area 38°42′5″N,115°9′13″E 2017-03-27(86), 2017-04-12(102), 2017-05-14(134)
黑龙江研究区 Heilongjiang research area 46°13′16″N,126°43′27″E 2017-07-07(188), 2017-09-09(252), 2017-09-25(268)
新疆研究区 Xinjiang research area 38°29′55″N,77°21′14″E 2017-06-07(158), 2017-06-23(174), 2017-08-10(222)

表2

MODIS与Landsat-8波段设置"

Landsat-8 OLI波段
Landsat-8 OLI band		 波长范围
Wavelength range
(nm)		 空间分辨率
Spatial resolution
(m)		 MCD43A4波段
MCD43A4 band		 波长范围
Wavelength range
(nm)		 空间分辨率
Spatial resolution
(m)
2 450-515 30 3 459-479 500
3 525-600 30 4 545-565 500
4 630-680 30 1 620-670 250
5 845-885 30 2 841-876 250
6 1560-1660 30 6 1628-1652 500
7 2100-2300 30 7 2105-2155 500

图2

技术流程框图"

图3

原始Landsat-8 OLI影像与ESTARFM融合后影像对比 A、D和G: 河北、黑龙江和新疆农区真实Landsat影像; C、F和I: 河北、黑龙江和新疆农区经ESTARFM预测后影像; B、E和H: 真实影像与预测影像细节表征对比图。"

图4

真实Landsat影像(横坐标)与ESTARFM融合后影像(纵坐标)反射率值对比 A1~A6: 河北农区真实Landsat影像与经ESTARFM模型融合后影像各波段的相关性; B1~B6: 黑龙江农区真实Landsat影像与经ESTARFM模型融合后影像波段的相关性; C1~C6: 新疆农区真实Landsat影像与经ESTARFM模型融合后影像波段的相关性。"

图5

真实影像与融合后影像NDVI结果比较 A: 河北农区(2017-04-12); B: 黑龙江农区(2017-09-09); C: 新疆农区(2017-06-23)。"

表3

原始Landsat-8 OLI影像与ESTARFM融合后结果相关性分析"

区域与影像获取日期
Region and acquisition date		 波段
Band		 决定系数
R2		 均方根误差
RMSE		 方差
Variance		 平均绝对偏差
MAD
河北Hebei Blue 0.8684 0.0451 0.0020 0.0355
2017-04-12 Green 0.8810 0.0566 0.0032 0.0447
Red 0.9446 0.0767 0.0059 0.0614
NIR 0.8551 0.1010 0.0102 0.0741
SWIR1 0.9309 0.1029 0.0106 0.0807
SWIR2 0.9081 0.1106 0.0122 0.0901
黑龙江Heilongjiang Blue 0.8913 0.0171 0.0003 0.0171
2017-09-09 Green 0.8906 0.0190 0.0004 0.0190
Red 0.9017 0.0252 0.0006 0.0252
NIR 0.7013 0.0278 0.0008 0.0278
SWIR1 0.7775 0.0332 0.0011 0.0332
SWIR2 0.7236 0.0443 0.0019 0.0443
新疆Xinjiang Blue 0.7842 0.0407 0.0017 0.0349
2017-06-23 Green 0.8084 0.0496 0.0024 0.0420
Red 0.8613 0.0679 0.0046 0.0594
NIR 0.8682 0.0932 0.0087 0.0765
SWIR1 0.8817 0.0836 0.0069 0.0589
SWIR2 0.9123 0.0973 0.0095 0.0839

图6

Landsat-8 OLI和MODIS数据的获取日期(DOY)"

图7

新疆阿克苏地区8 d时间间隔的高时空分辨率影像"

图8

融合后影像中各地物NDVI随时间变化情况 A: 植被; B: 沙漠; C: 建筑物; D: 水体。"

图9

农作物遥感分类结果"

图10

作物长势监测和分级图 A: 作物长势监测图; B: 棉花长势分级图; C: 水稻长势分级图。"

[1] Ju J, Roy D P . The availability of cloud-free Landsat ETM+ data over the conterminous United States and globally. Remot Sens Environ, 2008,112:1196-1211.
doi: 10.1016/j.rse.2007.08.011
[2] Pohl C, Genderen J L V . Review article multisensor image fusion in remote sensing: Concepts, Methods and Applications. Int J Remot Sens, 1998,19:823-854.
doi: 10.1080/014311698215748
[3] Jarihani A A, McVicar T R, Van Niel T G, Emelyanova I V, Callow J N, Johansen K . Blending Landsat and MODIS data to generate multispectral indices: a comparison of “Index-then- Blend” and “Blend-then-Index” approaches. Remot Sens, 2014,6:9213-9238.
doi: 10.3390/rs6109213
[4] 石月婵 . 基于多源遥感数据的高时空地表温度生成方法研究. 山东科技大学硕士学位论文, 山东青岛, 2013.
Shi Y C . Research on the Generation of High Temporal and Spatial land Surface Temperature Based on Multi-source Remote Sensing Data. MS Thesis of Shandong University of Science and Technology, Qingdao, Shandong, China, 2013 (in Chinese with English abstract).
[5] Gao F, Masek J, Schwaller M, Hall F . On the blending of the Landsat and MODIS surface reflectance: predicting daily Landsat surface reflectance. IEEE Transact Geosci Remot Sens, 2006,44:2207-2218.
doi: 10.1109/TGRS.2006.872081
[6] Knauer K, Gessner U, Fensholt R, Kuenzer C . An ESTARFM fusion framework for the generation of large-scaletime series in cloud-prone and heterogeneous landscapes. Remot Sens, 2016,8:425.
doi: 10.3390/rs8050425
[7] Hilker T, Wulder M A, Coops N C, Linke J, Mcdermid G, Masek J G, Gao F, White J C . A new data fusion model for high spatial-and temporal-resolution mapping of forest disturbance based on Landsat and MODIS. Remot Sens Environ, 2009,113:1613-1627.
doi: 10.1016/j.rse.2009.03.007
[8] Zhu X L, Chen J, Gao F, Chen X H, Masek J G . An enhanced spatial and temporal adaptive reflectance fusion model for complex heterogeneous regions. Remot Sens Environ, 2010,114:2610-2623.
doi: 10.1016/j.rse.2010.05.032
[9] Emelyanova I V, Mcvicar T R, Van Niel T G, Li L T, van Dijk A I J M . Assessing the accuracy of blending Landsat-MODIS surface reflectances in two landscapes with contrasting spatial and temporal dynamics: A framework for algorithm selection. Remot Sens Environ, 2013,133:193-209.
doi: 10.1016/j.rse.2013.02.007
[10] Chen B, Huang B, Xu B . Comparison of spatiotemporal fusion models: a review. Remot Sens, 2015,7:1798-1835.
doi: 10.3390/rs70201798
[11] Liao C H, Wang J F, Pritchard I, Liu J G, Shang J L . A spatio-temporal data fusion model for generating NDVI time series in heterogeneous regions. Remot Sens, 2017,9:1125.
doi: 10.3390/rs9111125
[12] Knauer K, Gessner U, Fensholt R, Forkuor G, Kuenzer C . Monitoring agricultural expansion in Burkina Faso over 14 years with 30 m resolution time series: the role of population growth and implications for the environment. Remot Sens, 2017,9:132.
doi: 10.3390/rs9020132
[13] Wu M Q, Niu Z, Wang C Y, Wu C Y, Wang L . Use of MODIS and Landsat time series data to generate high-resolution temporal synthetic Landsat data using a spatial and temporal reflectance fusion model. J Appl Remot Sens, 2012,6:063507.
doi: 10.1117/1.JRS.6.063507
[14] 石月婵, 杨贵军, 李鑫川, 宋健, 王纪华, 王锦地 . 融合多源遥感数据生成高时空分辨率数据的方法对比. 红外与毫米波学报, 2015,34:92-99.
Shi Y C, Yang G J, Li X C, Song J, Wang J H, Wang J D . Comparison of methods for generating high temporal and spatial resolution data by using multi-source remote sensing data. J Infrared Millimeter Wave, 2015,34:92-99 (in Chinese with English abstract).
[15] Wu M Q, Wu C Y, Huang W J, Niu Z, Wang C Y, Li W, Hao P Y . An improved high spatial and temporal data fusion approach for combining Landsat and MODIS data to generate daily synthetic Landsat imagery. Inform Fusion, 2016,31:14-25.
doi: 10.1016/j.inffus.2015.12.005
[16] 周巍, 朱荣, 张锦水 . 遥感辅助的农作物种植面积小域估计方法研究. 统计研究, 2015,32(7):81-86.
Zhou W, Zhu R, Zhang J S . A study on small area estimation for crop acreage in remote sensing assisted crop survey. Stat Res, 2015,32(7):81-86 (in Chinese with English abstract).
[17] 侯博展, 陈蜀江, 侯建楠, 黄铁成, 贾翔 . 基于多源多时相数据集的喀什市近70a城区空间扩展遥感研究. 冰川冻土, 2016,38:836-844.
Hou B Z, Chen S J, Hou J N, Huang T C, Jia X . Remote sensing study of urban spatial expansion in Kashgar City based on multisource and multi-temporal datasets in the last 70 years. J Glaciology Geocryo, 2016,38:836-844 (in Chinese with English abstract).
[18] 李国砚, 张仲元, 郑艳芬, 刘晓玫 . MODIS影像的大气校正及在太湖蓝藻监测中的应用. 湖泊科学, 2008,20:160-166.
Li G Y, Zhang Z Y, Zheng Y F, Liu X M . Atmospheric correction of MODIS images and its application in cyanobacteria bloom monitoring in Lake Taihu. J Lake Sci, 2008,20:160-166 (in Chinese with English abstract).
[19] Bai L L, Cai J B, Liu Y, Chen H, Zhang B Z, Huang L X . Responses of field evapotranspiration to the changes of cropping pattern and groundwater depth in large irrigation district of Yellow River basin. Agric Water Manage, 2017,188:1-11.
doi: 10.1016/j.agwat.2017.03.028
[20] Huang C, Chen Y, Zhang S Q, Li L Y, Shi K F, Liu R . Surface water mapping from Suomi NPP-VIIRS imagery at 30 m resolution via blending with Landsat data. Remot Sens, 2016,8:631.
doi: 10.3390/rs8080631
[21] 郭文静, 李爱农, 赵志强, 王继燕 . 基于AVHRR和TM的时间序列较高分辨率NDVI数据集重构方法. 遥感技术与应用, 2015,30:267-276.
Guo W J, Li A N, Zhao Z Q, Wang J Y . Constructing the time-series NDVI dataset with a high spatial and temporal resolution through fusing AVHRR with TM data. Remot Sens Tech Appl, 2015,30:267-276.
[22] 郝贵斌, 吴波, 张立福, 付东杰, 李瑶 . ESTARFM模型在西藏色林错湖面积时空变化中的应用分析(1976-2014年). 地球信息科学学报, 2016,18:833-846.
Hao G B, Wu B, Zhang L F, Fu D J, Li Y . Temporal and spatial variation analysis of the area of Siling Co lake in Tibet based on ESTARFM (1976-2014). J Geo-inform Sci, 2016,18:833-846. (in Chinese with English abstract).
[23] 杨贵军, 孙晨红, 历华 . 黑河流域ASTER与MODIS融合生成高分辨率地表温度的验证. 农业工程学报, 2015,31(6):193-200.
Yang G J, Sun C H, Li H . Verification of high-resolution surface temperature by fusion of ASTER and MODIS in Heihe River Basin. Trans CSAE, 2015,31(6):193-200 (in Chinese with English abstract).
[24] 吴炳方, 张峰, 刘成林, 张磊, 罗治敏 . 农作物长势综合遥感监测方法. 遥感学报, 2004,8:498-514.
Wu B F, Zhang F, Liu C L, Zhang L, Luo Z M . An integrated method for crop condition monitoring. J Remot Sens, 2004,8:498-514 (in Chinese with English abstract).
[25] Liu M X, Liu X N, Wu L, Zou X Y, Jiang T, Zhao B Y . A modified spatiotemporal fusion algorithm using phenological information for predicting reflectance of paddy rice in southern china. Remot Sens, 2018,10:772.
doi: 10.3390/rs10050772
[26] Zhang W, Li A, Jin H, Bian J H, Zhang Z J, Lei G B, Qin Z H, Huang C Q . An enhanced spatial and temporal data fusion model for fusing Landsat and MODIS surface reflectance to generate high temporal Landsat-like data. Remot Sens, 2013,5:5346-5368.
doi: 10.3390/rs5105346
[27] Gärtner P, Förster M, Kleinschmit B . The benefit of synthetically generated RapidEye and Landsat 8 data fusion time series for riparian forest disturbance monitoring. Remot Sens Environ, 2016,177:237-247.
doi: 10.1016/j.rse.2016.01.028
[28] Yang G J, Weng Q H, Pu R L, Gao F, Sun C H, Li H, Zhao C J . Evaluation of ASTER-like daily land surface temperature by fusing ASTER and MODIS data during the HiWATER-MUSOEXE. Remot Sens, 2016,8:75.
doi: 10.3390/rs8010075
[29] Fu D J, Chen B Z, Wang J, Zhu X L, Hilker T . An improved image fusion approach based on enhanced spatial and temporal the adaptive reflectance fusion model. Remot Sens, 2013,5:6346-6360.
doi: 10.3390/rs5126346
[1] 王磊，白由路，卢艳丽，王贺，杨俐苹. 基于GreenSeeker的冬小麦NDVI分析与产量估算[J]. 作物学报, 2012, 38(04): 747-753.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
京ICP备15008789号-2